Method and apparatus for achieving fault tolerance in packet switching systems with inverse multiplexing

摘要

A preferred embodiment incorporates a core switching network including a plurality of input and output ports and a plurality of switching paths between the input and output ports. Preferably, a plurality of input controllers and output controllers arc provided, with each of the input controllers being attached to a plurality of the input ports and each of the output controllers being attached to a plurality of the output ports. The core switching network also preferably is configured to receive a first flow of data packets from one of the input controllers and deliver a second flow of data packets to one of the output controllers, and the packet switching system is configured to restore the second flow of data packets despite a loss of information from at least one of the data packets upstream of the one of the output controllers so that the second flow has an order of remaining ones of the data packets corresponding to an order of the data packets in the first flow. Additionally, methods and apparatuses for checkpointing inverse multiplexed network connection traffic for recovering synchronization of the data are provided.